Re: *scratch* lost

["B. T. Raven" <nihil@nihilo.net>]

Pascal J. Bourguignon wrote:
> "B. T. Raven" <nihil@nihilo.net> writes:
> 
>> I was doing some key mapping in .emacs and instead of saving the file,
>> restarting Emacs, and testing the new assignment I was doing M-x eval
>> region on changed areas of .emacs Somehow all (or many) keychords
>> started acting strangely, e.g. C-x 1 would open subr.el in a new
>> buffer instead of showing current buffer alone in frame, even cursor
>> movement keys did something else. I killed .emacs buffer without
>> saving (losing changes) and then exited Emacs. On restarting
>> everything looks and works normally exept that *scratch* is empty
>> (three commented lines are missing) and it is in Fundamental mode
>> instead of Lisp Interaction. Can I get the old *scratch* buffer back?
>> How?
> 
> (defun make-scratch ()
>   (interactive)
>   (if (get-buffer "*scratch*")
>      (switch-to-buffer (get-buffer "*scratch*"))
>      (progn
>         (switch-to-buffer (get-buffer-create "*scratch*"))
>         (insert ";; This buffer is for notes you don't want to save, and for Lisp evaluation.
> ;; If you want to create a file, visit that file with C-x C-f,
> ;; then enter the text in that file's own buffer.
> 
> ")
>     (emacs-lisp-mode))))
> 
> 
> M-x make-scratch RET
> 
> 
> 
Thanks, Pascal. I made the mode for *scratch* (lisp-interaction-mode) so 
that C-j alone evaluates but if I needed your function to get a normal 
*scratch* every time I would have reinstalled Emacs instead. My .emacs 
(loaded under switch --debug-init) didn't seem to have any problems 
except for an empty *scratch* in fundamental mode but a normal *scratch* 
was produced with emacs -Q. Somehow a customization of org-agenda-file 
(maybe with extra quotes or something but whatever it was didn't hang up 
debug-init; I didn't pay close enough attention) was producing an empty 
*scratch* in fundamental mode. ??? Anyway I saved your function for 
later use if necessary.
Now,  '(org-agenda-files (quote ("c:/emacs/org"))) under 
custom-set-variables doesn't cause a problem.

Ed